10 Reasons to try Chameleon Free

Do we really need to give a reason, other than "it's free"!

1. Well, as we said above, it's free! Genuinely free, no credit cards, no commitments, no tie-ins. We only ask for your email address so we can get you licensed for a free version.

2. You will be able to generate data that is both free, as in at no cost to yourself, but also, and perhaps more importantly, free of any IPR, GDPR, and privacy concerns; can you say that about the rest of your data?

3. Even a small amount of additional data can improve model performance, if it is the right data! Chameleon's workflow encourages a new paradigm for ML model training. Guided by failure cases, you can create new, novel data that will help the model repair its flaws. And if the first additional data doesn't work, no problem, you can easily create some more.

4. It's fun! Using Chameleon, being the Director, producer and star is a lot of fun. If data creation could be gamified, this would be the game! The unique UI makes it fun and easy to create new data.

5. You may actually improve your skills as an ML engineer/Data Scientist/Product Manager. Giving more thought to the data part of the ML process will help you better optimize your own workflows to get better results, faster.

6. Ever considered how your model would perform if the inferencing camera was placed in a different position? Or in different lighting conditions? Or with a population with different skin tones? Why not use the free version of Chameleon to create some data to test and stretch your model to breaking point?

7. Do some competitive analysis. Try and see how quick and easy it is to gather a dataset of 1,000 different people, with different skin tones, in different clothing with different backgrounds, from different camera angles with different lighting conditions with your current solution? Chameleon's digital humans, passes, generators and cloud cameras automate the process, resulting in varied datasets produced in minutes, not hours or days.

8. Use the capability of multi camera to try some new algorithm development that has not been possible for you before due to lack of synchronized, perfectly annotated multi-camera data.

9. Try a new annotation type, or a mix of annotations to supercharge your algorithm? Have you considered using accurate skeleton keypoints? 3D bounding boxes? Velocity information? Your creativity can be inspired by making use of these advanced and accurate annotation types to improve your algorithm accuracy and robustness.

10. Did I mention the fact that it's free? Just checking!